This is a wheel made with compliant mechanisms that can traverse obstacles up to 70% of its height and allow robots to navigate complex terrain.

This model must be printed with TPU (like 95A), you might have barely some luck with PETG but probably it will be too stiff and break.

I have included a 3d model of the wheel with diameter 120mm and thickness 10mm, with a hole for a 6mm drive shaft and 6 holes spaced at 31mm, it's finetuned for light applications, if you want to modify the model i have included the inventor CAD file. There is also a blank wheel model with no driveshaft holes so that you can make the holes yourself in bambu studio with negative modifiers

For heavier loads i suggest starting with scaling the model vertically, if that isn't enough you can increase the strenght by opening the inventor file and 

• increasing the thickness
• decreasing the “elasticity finetune” size (it's the size of the hole in the joints wich makes them more flexible) 
• decreasing the size of the lightening holes.

If this models gains some traction i will revision it further, there is still some margin for improvement.

If you print this on a PEI bed i suggest you use some stuff like glue because tpu sticks to the bed a lot and this model is very hard to remove.

The credits for this invention go to this paper: (or whatch the youtube video)

T. Godden, B. W. Mulvey, E. Redgrave and T. Nanayakkara, "PaTS-Wheel: A Passively-Transformable Single-Part Wheel for Mobile Robot Navigation on Unstructured Terrain," in IEEE Robotics and Automation Letters, vol. 9, no. 6, pp. 5512-5519, June 2024, doi: 10.1109/LRA.2024.3389828.

Abstract: Most mobile robots use wheels that perform well on even and structured ground, like in factories and warehouses. However, they face challenges traversing unstructured terrain such as stepped obstacles. This letter presents the design and testing of the PaTS-Wheel: a Passively-Transformable Single-part Wheel that can transform to render hooks when presented with obstacles. The passive rendering of this useful morphological feature is guided purely by the geometry of the obstacle. The energy consumption and vibrational profile of the PaTS-Wheel on flat ground is comparable to a standard wheel of the same size. In addition, our novel wheel design was tested traversing different terrains with stepped obstacles of incremental heights. The PaTS-Wheel achieved 100% success rate at traversing stepped obstacles with heights ≈70% its diameter, higher than the results obtained for an equivalent wheel (≈25% its diameter) and an equivalent wheg (≈61% its diameter). This achieves the design objectives of combining the energy efficiency and ride smoothness of wheels with the obstacle traversal capabilities of legged robots, all without requiring any sensors, actuators, or controllers